DOCUNENT RESUME

ED 343 141 CS 213 226

AUTHOR Haswell, Richard H.; Tedesco, Janis E.

TITLE Gender and the Evaluation of Writing.

PUB DATE Nov 91

NOTE 14p.; Paper presented at the Annual Meeting of theNational Council of Teachers of English (Slst,Seattle, WA, November 22-27, 1991).

PUB TYPE Reports - Swsearch/Technical (143) --Speeches'Ionference Papers (150)

EDRS PRICE MF01/PC01 Plus Postage.

DESCRIPTORS Attitude Measures; Cultural Influences; Females;*Freshman Composition; Higher Education; Males; *SexDifferences; ltereotypes; *Writing Evaluation;

Writing ResearchIDENTIFIERS Text Factors

ABSTRACTA study examined the effects of gender-linked

features of writing upon raters' judgments about writing quality.

Sixty-four subjects were interviewed: 32 teachers and 32 freshmancomposition students. Subjects were asked to evaluate two essays, onewritten by a woman and one by a man. In interviews, the sabjects were

asked to: (1) offer suggestions for revision; (2) describe strong andweak features of the essay; (3) identify the one most important piece

of advice they would offer to tNa writer; (4) rank the merit of the

essay on a scale of one to five, with one being low and five high;

(5) guess the gender of the writer, and identifying clues in the textthat suggested gender. The writer's gender had no influence on the

success rate in gueswing gender, and each group's guesses were often

wrong by a ratio of two to one. Both textual and nontextual clues to

gender were identified. Male writers offered fewer gender clues in

their essays, but tended to offer more textual than nontexi.aal clues.

Gender did influence how essays were rated. Results suggest thatthere is a culturally determined way of looking at gender thatstudents and teacher: bring to the evaluation process. (Four figures

are included; two appendixes containing students' texts areattached.) (SG)

*****************************************M*****************************

* Reproduction* supplied by EDRS are the best that can be madefrom the original document.

.?

"PERMISSION TO REPRODUCE THIS

MATERIAL HS BEEN GRANTED BY

TO THE EDUCATIONAL RESOURCES

INFORMATIONCENTER (ERIC)."

GENDER AND THE EVALUATION OFWRITING

Richard H. Haswell and Janis E. Tedesco

Department of EnglishWashington State University

Paper presented at theNCTE National Convention(Seattle, November 1991)

us. Deragraeasar OF EDUCATIONOffee Educaouno Reeeetch end aroprovernentEOUCAnortu RESOURCES latrOAPAATION

CE NYE R ER1C)IlVirle Amur/Iwo has aeon releoduce eawaved f Va. MP petiOn 131 ergot...m.0nougmahrog

I ' Wn o. changes have Wee mode In smiuoveseemauchon gimlet

Pooh?, of velvet oppugns meted emus documem do nos novreuerdv Attatrgent offk.stOERI Dovaon ocescr

What is the relationship between writing instruction and gender? More

specifically, is gender a factor in the way teachers evaluate of writing?*

Our research generates data suggesting that there is a degree to which

gender-linked features of a text affect raters' judgments about the quality

of the writing. Our data also uncover the :mportance of the researcher-

interviewer's gender in relation to the kinds of qualitative judgments

offered by research subjects. If we adopt the current understanding of

gender and defme it as a psychological and social construction, then we

have found that teacher evaluation can and does interact with gender and so

contributes, positively or negatively, in that construction. The present

paper is the first, preliminary report of our research into the area. At

present the project is still in progress.

In the fall of 1991, we interviewed 64 randomly selected subjects: 32

teachers with varying levels of experience (new teaching assistants and

* This paper does not attempt to summarize previous research dealing with gender

and tewher evaluation. Generally, other saidies (such as those conducted by Duane

Roen, Shirley IC. Rose, Donald Rubin, Kathryn Greene and Judith Barnes) either

have a small number of subject or take a case-study approach. There is clean

for broadbased, empirical studies such as ours.

IEST con AVAILABLE

seasoned instructors) and 32 entering college freshman in regular

composition (English 101). These groups and subgroups were evenly

divided by sex. We asked them to evaluate two essays on a blind reading

basis, one written by a woman and one by a man, from an English 101

class, spring semester 1989 (see appendix A for Victoria's text, labeled

#18, and appendix B for Kevin's text, labeled #26). The two impromptus

were written in class, as twenty-minute responses to the following prompt:

"How would you describe your 'search for truth' and the process you use

to pursue it?"

We taped half-hour interviews with these 64 subjects, asking them in a

conference-like setting to imagine themselves in a teaching or peer editing

situation wherein they might offer the student-writer suggestions for

revision. We developed an interview protocol that asked our subjects 1) to

offer suggestions for revision, 2) to describe strong and weak features of

the essay, 3) to identify the one most important piece of advice they would

offer the writer, 4) to rank the merit of the essay on a scale of one to five,

with one being low and five high, and 5) to guess the gender of the writer,

identifying clues in the text that suggested either a male or female. We

divided our teachers and students by gender, each of us interviewing an

equal number of male and female readers. We also alternated which essay

was read first, choosing to identify the gender of the second writer so that

our data would reflect evaluation in two different situations: when the

gender of the writer is known and when the gender is unknown to the

reader.

We are now processing over 588 pages of typed transcriptions,

classifying both pedagogical comments and gender clues. Although we have

- 23

yet to run tests of statistical inference, the data has already yielded several

evident trends that we can descrIbe as follows:

(1) Our readers responded to gender even on a blind reading basis.

Figure #1 shows that each group's gender guesses or ascriptions (the groups

being female and male students, and female and male teachers) were often

wrong by a ratio of 2:1. This is not surprising, considering that we selected

two essays that we thought would be difficult in terms of gender

assignment. What is significant is that only 2 out of 64 readers could not

formulate any kind of gender ascription. Indeed, forty out of 64 readers

had developed on their own initiative a sense of the writer's gender before

they were asked to make any gender identification. Figure #2 compares

correct and incorrect ascriptions while linking the gender of the reader to

the gender of the writer. The frndings show that the gender of the writer

had no influence on the success rate of ascription for either female or male

readers.

(2) When subjects were asked to locate gender clues, two types

surfacedtextual, when the reader identifies specific written expressions as

suggestive of gender, and nontextusL1,when the reader relies on more

general gender patterns extending beyond the text (these types of clues

were often stereotypical, like "women are better writers," "women are

more emotional," "men are more logical and analytic," "men are unwilling

to talk about emotion," etc). There were several interesting trends, as

Figure #3 illustrates. First, male readers produce significantly fewer gen-

der clues with both essays. Second, male readers offer more textual than

non-textual clues. Third, female readers were most willing to offer textual

clues when reading the essay written by a female. We also found that

- 3

4

three out of five clues offered by our readers addressed characteristics of

men and of masculine writing.

When we marked all the content or language features in the essays that

had suggested the writer's gender to our readers, we found that nearly 60%

of the texts was used. This result is especially significant, since all but a few

of our readers expressed the belief that gender should not play a part in

teacher evaluation. What is also significant is that with Victoria's essay,

which most of our readers thought yas written by a man, there were ample

textual clues identified as "feminine." With Kevin's essay, which most of

our readers thought was written by a woman, there were ample textual

clues identified as "masculine." Often the same content and language

features in essay #18 and essay #26 were identified as being masculine clues

by some readers and feminine clues by others. In such cases, the selected

features were described differently. When feminine gender clues were

found (by both male and female readers), Victoria's essay is described in

the following way:

This writer thinks in tenns of tfontext. She would defy thelaw in order to protect people she cares about. Shequalifies her attitude stoat the process slw uses to searchfor truth with the admission that there are drawbacks andthat sometimes she makes mistakes. She clearly valuesmoral issues. She is open to emotion and relies on her owninstinct. Although she can be hasty at times, she iscomfortable in looking inside herself for answers. Heressay is thoughtful, well organized, and contains fewgrammatical and syntactic flaws. She shows some lapses inthe formr style, using slang like "gut instinct" and "loadof bull."

- 4 -

When masculine clues were found (by both female and male readers), the

same essay is described in other terms:

This writer is decisive, logical and linear in his thinking.He hates to be proven wrong. He's corrpetitive and isfundamentally self-reliant and independent even though hewill weigh evidence from other sources. He is assuredexcept when it comes to emotions. He has a vigorous stylethat is straightforward and aggressive, as evident byphrases like "gut instinct" and "load of bull."

(3) Even though our readers insisted that evaluation should be gender-

neutral, our research indicates that gender did influence how the essays

were rated (see Figure #4). Victoria's essay was rated higher by both male

and female readers, whether the gender was known or unknown. But in

comparing the ratings of each essay when the gender was first known and

then unknown, results indicate that 1) female readers gave Victoria's essay

a lower score when they knew it was written by a female; 2) female

readers gave Kevin's essay a higher score when they knew is was written

by a male, 3) male readers gave Victoria's essay a higher score when they

knew the gender, and 4) male readers gave Kevin's essay a lower score

when they knew the gender.

We are currently analyzing patterns in terms of the gender of the

interviewer and in terms of the relation of the gender clues to the kinds of

suggestions for revisions offered by our readers. But the results we have

just described, however incomplete and tentative, demonstrate that our

teachers and students did create a picture of the writer in terms of gender,

sometimes incorrectly, often unconsciously, and almost always in

contradiction to their implicit denial that readers should attend to such

features. This indicates that gender does surface in the text readers look

- 5,

for clues (and they do), they can fmd them. Further, we have also learned

that readers respond differently to student writing, depending on whether

they know or even think that the text is written by a male or female. The

image of the writer as male or female produces a different sense of overall

relative value and evokes a different vocabulary with which the readers

describe the essays. This last finding, of course, has major implications for

the teaching and assessment of writing.

Our results suggest that there is a culturally determined way of looking

at gender that both writing students and composition teachers bring to the

evaluation process. Deborah Cameron and Cherie Kramerae might call this

behavior "folk linguistics" and Judith Barnes might call it "gender

framing." It clearly operates both for our male and female readers as they

identified textual and non-textual gender clues. As we continue our study,

one of the patterns we will be looking at more closely is the evolution of

folk linguistics in terms of the impact of feminist criticism and women's

studies. Are such advances in the curriculum of secondary and

postsecondary education eliminating an already existing system of

assumptions, or simply substituting a new one, with perhaps a new

vocabulary and an alternative writing style and genre that are now to be

privileged?

The majority of our readers emphasizes the need for person-centered

assessment during writing conferences. In this context, cw r readers

acknowledged that gender may be a factor in student/teacher interaction.

They suggested that women students may be more vulnerable or sensitive,

while men might be more impatient and just want to know how to "fix" the

essay for a higher grade. But readers resisted the idea that gender may be a

feature in the text itself or a factor in their evaluation of writing. As we

- 6 -.11

have said, our initial findings argue that they are self-deceived, at least in

part.

By implication, the preliminary stages of our analysis entertains the

possibility that a writing teacher cannot be gender-neutral and value-free at

the same time. A teacher can be consciously open and sensitive to a student

writer and yet force that writer into adopting practices that reflect

unexamined assumptions and judgments. In tryhig to efface gender, we are

in fact eliminating the presence of the author in his or her textperhaps

the ultimate and most violent form of exclusion that can occur in a

composition classroom. In attempts to be gender neutral, we are de-

authoring the student text by degendering the student's voice.

A postscript. At the end of our research, we interviewed Victoria and

Kevin, the writers of these two essays. We asked them to reread their

responses and then invited their evaluation. Especially Victoria's comments

supported our conclusions. Victoria felt that her old essay still reflected

her search for truth. When we told her that some readers had trouble

identifying her gender, she noted the phrases that to her captured her

femininity, phrases like "this process involves a gut instinct" and "it's up to

me to figure out if somebody/source is feeding me a load of bull"those

very words that for most of our readers indicated a male writer. These

phrases were also often noted as unacceptable for academic discourse; they

were too informal and too imprecise. But for Victoria, they mattered. "It's

me," she told us. "It says everything I stand for . . . the voice is all me."

- 7 -s. a..

15

10

5

o

-

_

FIGURE 1

1 maw corrFEMME STUDENTS

ncorr corrFENALE TEMERS

1 neorr corrnilLE STUDENTS

Ii ncorr corrnag TEACHERS

CORRECTNESS OF GENDER ASCRIPTION

- 8 _

9

FIGURE 2

15 -

cam 1w sterr I momWILE ESSAY : FEMALE READER tynnu MAW : RALE READER

NILE ESSRY : FEMALE REAMER ME ESSAY : 11RU RIMER

CORRECTNESS OF GENDER ASCRIPTION

- 9

10

4

.. FIGURE 3

2.50

2.00

1.50

1. 00

0.50

0.00Female Male

MALE AND FastALE READERS

I.

Textual Cluesfor Essay 18

Non-textual cluesfor Essay 18

Textual duesfor Essay 26

Non-textual cluesfor Essay 26

TEXTUAL AND NON.TEXTUAL GENDER CLUES

Se

FIGURE 4

3.3

2.1111

2.1111SCALL.26

I SCALE_ IStea km& male taws es Urea sale km*FEMALE READERS MALE READERS

QUALITY RATINGS WHEN GENDER OF AUTHOR

13 KNOWN AND UNKNOWN

12

a

[Victoria's text]

APPENDIX A

18

The process by which I search for "truth" is dependent upon what kind

of an answer I am looking for.

For example, if I were looking for the answer to a question of

morality, I would look within myself. I believe that only I can know

If what I am doing or what I am saying is "good" or "bad". / use

myself and my own personal values to determine the difference between

right and wrong. I use the beliefs I hold strongly to act as a kind of

guide to help me through some more complex moral decisions. Forinstance, I believe in obeying the law, but I realize that the law Isonly es perfect as those who made it. Thus, if on occasion ariseswhore someone is in danger or is hurt and helping them would conflictwith the law. I would tend to ignore that specific law.

If I were searching for au answer to a question involving knowledge, I

would first look to myself and see how much I know about the particularsubject or question / am contemplaLing. I then will tike whatknowledge I have an compare it to what other people (or other sources)

know. This process also involves a gut instinct, for I'm the only onewho can decide if a source or a person Is giving me a qualifiedanswer. In other words, it's up to me to figure out if somebody/source

is feeding me a load of bull. Once I have the chance to gather as muchinformation that I can, I will try to make as accurate answer aspossible. It should be noted that on some occasions I choose not to usother people/sources to find the truth. Sometimes I am able to find

the answers without the help of anyone else.

In conclusion I would like to say that, while these methods for finding

my own kind of truth seem to work fairly well, I realize that there re

drawbacks. One involves emotion. Sometimes, in cases where there is alot of emotion going on, I am apt to make decisions that are too

hasty. Another drawback is the amount of time I have to make these

decisions. In cases math as these, I just go with what I knowdefinitely and my instinct. Also, like any other person, I don't like

to be proven wrong, but I guess lea something I've learned to live

with.

- 12 -

13

[Kevin's text]

APPENDIX B

26

When Plato describes a person's "search for truth", he was the

"allegory of the cave". Bow would you describe your "search for truth"

and the process you use to pursue it/

When I find myself searching for truth I usually try to find it in

friends and my family. I also find it through my own self, because I

have to take in the information my friends and parents give se and

decide what I want to believe is real. So, I basically decide what isreal through my own self and my own beliefs, but I get most of the

information from other people outside myself.

To find truth la something that.comes naturally to me I guess. When I

take In information that my friends or my family is telling me I have

to take in all the good, truthful information and through out all the

bad information. Something that they believe is truthful may not be

truthful to me. I am my own person and I like to make my own decisions

so when I get the information I take all the variables that go along

with it to make sure my decision will be right. There are so manythings that could influence my decision, but the biggest thing is

whether I trust the source I am getting my information from. That is

why when people I do not know try to give ne information I really don't

pay attention. I mean I pay attention because I am interested, but Iam not going to take what they are saying as truthful. Only if I

thought that it could be truthful would I then go to a friend or family

and ask them to elaborate on the subject that I brought up. So, to me,

all truth is something that I have to find myself through others. To

know if somebody's information Is really true or false Is my own

decision. I have to think whether I believe the information is real or

true. In this part of the decision making, everything comes down on my

own decision. This Is the hardest part, trying to decide whet is true

and what is false. I see it as what I believe in and what I want to

see is real, is real. Sven if everybody else sees the same thing 48false and I vent to believe it is real, it will be real. This is the

one problem with my decision making process on what is real and whatisn't real because if it happens that the information that I believe is

real is not real, by definition, then I go all through my life

believing it is real. This is why I have to take so much caution andtime to make the right decision on what is real, who do I get theinformation from, and making the final decision.

The information that I get from other people than myself is when I get

the information to decide what is real and what is the truth. Making

the right dcision I encounter lots of variables, but I have to make

the right choice because it stays with se my whole life. The

information comes from the outside by the truth comes from my inside.

- 13 -